[ci-fix-net11] Needs review: de-flake ActiveWindowTracker Windows device tests — bounded condition wait instead of fixed 100 ms delay (refs #36 [Content truncated due to length]

[github-actions]

Note

Are you waiting for the changes in this PR to be merged?
It would be very helpful if you could test the resulting artifacts from this PR and let us know in a comment if this change resolves your issue. Thank you!

Target branch: net11.0
Refs: #36043
Attempt: 1/5

Summary

The ActiveWindowTracker Windows device tests intermittently fail with an empty-collection assertion — e.g. SwitchingWindowsPostsToTheNewWindow failing as Assert.Contains() Failure ... Collection: []. This is a test-quality flake (Step 4.7 bucket b), not a product defect.

Root cause

BaseWindowMessageManager_Tests.PostTestMessageAsync posts an asynchronous window message via PostMessage and then waits a fixed await Task.Delay(100) before the test asserts the message was received. PostMessage only delivers once the message pump dispatches it; under CI load that can exceed 100 ms, so the positive assertions occasionally run before the message arrives and observe an empty collection.

Observed intermittently (green-on-retry) in 2 of the last ~10 net11.0 maui-pr-devicetests builds (1473153, 1455596) — the classic flaky signal.

Fix (de-flake — not a mute)

Replace the fixed-delay-then-assert pattern with a bounded condition wait (WaitForMessageAsync) for the three positive-assertion tests:

• ActivatedBeforeStartRecievesMessage
• StartBeforeActivatedRecievesMessage
• SwitchingWindowsPostsToTheNewWindow

The helper polls (yielding to the message pump) until the expected message is observed, up to a generous 5 s ceiling. It returns as soon as the message arrives, so there is no artificial slowdown, and if the message is never delivered (a genuine regression) the wait times out and the unchanged Assert.Contains still fails. Negative-assertion tests (StoppedTrackerDoesNotTrack, SwitchingWindowsDoesNotPostToTheOldWindow) intentionally keep the fixed delay, since absence cannot be polled for.

Why this is not a mute

• No assertions removed or weakened — every original Assert.Contains / Assert.DoesNotContain is intact.
• No [Retry] / [Repeat] / [ActiveIssue] / [Ignore] / category exclusion / timeout-hiding change added.
• A real "message never delivered" regression still fails the test after the ceiling, so the change cannot hide a product bug.

Validation

Not runner-validated. These are Windows-only device tests (net11.0-windows, user32.dll PostMessage + Microsoft.UI.Xaml.Window) and this fixer runs on a Linux runner with no Windows workload, so the Essentials.DeviceTests project cannot be built or executed here. Please validate via a maui-pr-devicetests run on Windows (or run the Windows ActiveWindowTracker device tests locally) before merging.

Files changed

• src/Essentials/test/DeviceTests/Tests/Windows/BaseWindowMessageManager_Tests.cs — add the WaitForMessageAsync condition-wait helper.
• src/Essentials/test/DeviceTests/Tests/Windows/ActiveWindowTracker_Tests.cs — await the condition before the three positive assertions.

_{Generated by the CI Failure Fixer (net11.0-only). Target branch: net11.0 · Refs: #36043 · Attempt: 1/5}

Generated by CI Failure Fixer (net11.0) · 2.7K AIC · ⌖ 166.9 AIC · ⊞ 67K · ◷

[MauiBot]

AI Review Summary

@github-actions[bot] — new AI review results are available based on this last commit: eee09e3. To request a fresh review after new comments or commits, comment /review rerun.

---

🗂️ Review Sessions — click to expand

🚦 Gate — Test Before & After Fix

Gate Result: ⚠️ INCONCLUSIVE

Platform: ANDROID

⚠️ verify-tests-fail.ps1 exited before writing a verification report. Diagnostics below.

Exit code: 3

Auto-detected:

• Test filter: Windows_ActiveWindowTracker_Tests

Likely cause:

• Device/emulator setup failed (env error class).

Artifacts written before exit:

• test-failure-BaseWindowMessageManager_Tests.log (287 KB)
• test-failure-BaseWindowMessageManager_Tests.log.attempt2 (449.5 KB)
• test-failure-Windows_ActiveWindowTracker_Tests.log (400 KB)
• verification-log.txt (0.2 KB)

Gate output log (last 60 lines)

  Starting android Device/Emulator
═══════════════════════════════════════════════════════════
✓ Device ready: emulator-5554
═══════════════════════════════════════════════════════════
  Running Tests
═══════════════════════════════════════════════════════════
Running: dotnet xharness android test --app /home/vsts/work/1/s/artifacts/bin/Essentials.DeviceTests/Release/net11.0-android/com.microsoft.maui.essentials.devicetests-Signed.apk --package-name com.microsoft.maui.essentials.devicetests --device-id emulator-5554 -o artifacts/log --timeout 01:00:00 -v --arg TestFilter=BaseWindowMessageManager_Tests
Target:  android-emulator-64
Device:  emulator-5554
═══════════════════════════════════════════════════════════
  Test Results
═══════════════════════════════════════════════════════════
═══════════════════════════════════════════════════════════
═══════════════════════════════════════════════════════════
─────────────────────────────────────────────────
📱 Test 2/2: [DeviceTest] Windows_ActiveWindowTracker_Tests
🧪 Running device tests: Essentials on android
   Filter: Windows_ActiveWindowTracker_Tests
═══════════════════════════════════════════════════════════
  MAUI Device Tests Runner
═══════════════════════════════════════════════════════════
✓ xharness found: local dotnet tool
✓ dotnet found: /home/vsts/work/1/s/.dotnet/dotnet
Project:       Essentials
Project Path:  src/Essentials/test/DeviceTests/Essentials.DeviceTests.csproj
Platform:      android
Configuration: Release
Test Filter:   Windows_ActiveWindowTracker_Tests
═══════════════════════════════════════════════════════════
  Building Essentials Device Tests for android
═══════════════════════════════════════════════════════════
Running: dotnet build src/Essentials/test/DeviceTests/Essentials.DeviceTests.csproj -c Release -f net11.0-android /p:TreatWarningsAsErrors=false /p:AndroidPackageFormat=apk
✓ Build succeeded
✓ App found: /home/vsts/work/1/s/artifacts/bin/Essentials.DeviceTests/Release/net11.0-android/com.microsoft.maui.essentials.devicetests-Signed.apk
═══════════════════════════════════════════════════════════
  Starting android Device/Emulator
═══════════════════════════════════════════════════════════
✓ Device ready: emulator-5554
═══════════════════════════════════════════════════════════
  Running Tests
═══════════════════════════════════════════════════════════
Running: dotnet xharness android test --app /home/vsts/work/1/s/artifacts/bin/Essentials.DeviceTests/Release/net11.0-android/com.microsoft.maui.essentials.devicetests-Signed.apk --package-name com.microsoft.maui.essentials.devicetests --device-id emulator-5554 -o artifacts/log --timeout 01:00:00 -v --arg TestFilter=Windows_ActiveWindowTracker_Tests
Target:  android-emulator-64
Device:  emulator-5554
═══════════════════════════════════════════════════════════
  Test Results
═══════════════════════════════════════════════════════════
═══════════════════════════════════════════════════════════
═══════════════════════════════════════════════════════════
==========================================
VERIFICATION RESULTS
==========================================
  📱 [DeviceTest] BaseWindowMessageManager_Tests: PASSED ❌ (should fail!)
  📱 [DeviceTest] Windows_ActiveWindowTracker_Tests: PASSED ❌ (should fail!)
╔═══════════════════════════════════════════════════════════╗
║              VERIFICATION FAILED ❌                       ║
╠═══════════════════════════════════════════════════════════╣
║  2/2 test(s) PASSED but should FAIL!                   ║
║  Those tests don't reproduce the bug. Revise them!        ║
╚═══════════════════════════════════════════════════════════╝

---

📱 UI Tests — Essentials

Detected UI test categories: Essentials

⏭️ Deep UI tests — 0 passed, 0 failed across 1 category on platform-pool agent (replaces in-process counts above). 1 category reported 0 tests.

🧪 UI Test Execution Results (deep, platform pool)

Category	Tests	Snapshot diffs
Essentials	0 tests	—
📎 Download drop-deep-uitests artifact (TRX + snapshot diffs)

---

📋 Pre-Flight — Context & Validation

Issue: #36043 - [ci-scan-net11] Essentials.DeviceTests: SwitchingWindowsPostsToTheNewWindow fails on Windows — Assert.Contains collection empty (net11.0)
PR: #36092 - [ci-fix] Needs review: de-flake ActiveWindowTracker Windows device tests — bounded condition wait instead of fixed 100 ms delay
Platforms Affected: Windows device tests. Requested try-fix test platform: android, which cannot execute these Windows-only tests.
Files Changed: 0 implementation, 2 test

Key Findings

• The reported failure is in SwitchingWindowsPostsToTheNewWindow, where a positive assertion sometimes observes an empty collection after PostMessage plus a fixed 100 ms delay.
• PR #36092 keeps assertions intact and replaces fixed-delay positive checks with a bounded polling helper in the three positive ActiveWindowTracker tests.
• The changed tests are Windows-only (Tests/Windows, WinUI Window, user32.dll), so Android execution is not a meaningful regression gate.
• No prior PR review comments or inline review findings were available from public GitHub API data.
• GitHub CLI is unauthenticated in this environment; required-check and Azure DevOps test status could not be fully classified.

Code Review Summary

Verdict: NEEDS_DISCUSSION
Confidence: low
Errors: 0 | Warnings: 0 | Suggestions: 0

Key code review findings:

• No concrete code-level issues found in the PR diff.
• CI/gate status is undetermined because gh pr checks --required is unavailable and the prior gate result was already inconclusive.
• Blast radius is limited to Windows Essentials device tests; no product runtime or public API changes.

Fix Candidates

#	Source	Approach	Test Result	Files Changed	Notes
PR	PR #36092	Add WaitForMessageAsync(Func<bool>) polling and await expected message before positive assertions.	WARNING INCONCLUSIVE (prior gate could not build/run)	ActiveWindowTracker_Tests.cs, BaseWindowMessageManager_Tests.cs	Original PR fix; code review found no concrete issues.

---

🔬 Code Review — Deep Analysis

Code Review — PR #36092

Independent Assessment

What this changes: Adds a bounded polling helper for Windows PostMessage delivery and uses it in three positive ActiveWindowTracker device tests before asserting.
Inferred motivation: De-flake tests that assert before the Win32 message pump dispatches the posted message.

Reconciliation with PR Narrative

Author claims: Fixes intermittent Windows device-test failures from fixed 100 ms delay after async PostMessage.
Agreement/disagreement: Agrees. The code preserves assertions and waits only for positive delivery cases.

Prior Review Reconciliation

No prior error findings found.

Blast Radius Assessment

• Runs for all instances: no — test-only helper.
• Startup impact: no.
• Static/shared state: no.

CI Status

• Required-check result: gh pr checks --required unavailable: gh is unauthenticated.
• Public check-run result: maui-pr failed; failures include macOS AOT and Windows Helix Unit Tests Debug.
• Classification: undetermined.
• Action taken: ci-analysis unavailable; confidence capped low.

Findings

No Error, Warning, or Suggestion findings. Expert reviewer also reported no important findings.

Failure-Mode Probing

• Message arrives after 100 ms: helper yields and keeps polling until observed.
• Message never arrives: final Assert.Contains still fails after timeout.
• Negative tests: unchanged; absence assertions still use existing fixed delay.

Verdict: NEEDS_DISCUSSION

Confidence: low
Summary: Code change looks sound and localized, but CI is red/undetermined and required checks could not be verified with gh auth unavailable. Per skill rules, this cannot be LGTM until CI is classified or rerun.

---

🛠️ Fix — Analysis & Comparison

Fix Candidates

#	Source	Approach	Test Result	Files Changed	Notes
1	try-fix	TaskCompletionSource event signal from tracker.WindowMessage.	BLOCKED	2 files	More precise than polling, but more boilerplate and cleanup. Android cannot run Windows-only regression.
2	try-fix	Replace test PostMessage with synchronous SendMessage.	BLOCKED	2 files	Most deterministic and fastest, but lowers fidelity to queued message delivery. Android cannot run Windows-only regression.
3	try-fix	Await DispatcherQueue barrier turns after queued PostMessage.	BLOCKED	2 files	Preserves queued delivery but relies on Win32/DispatcherQueue ordering; not clearly stronger than polling. Android cannot run Windows-only regression.
4	try-fix	Add internal synthetic dispatch seam and test tracker state directly.	BLOCKED	3 files	Deterministic abstraction-boundary fix, but adds production test seam and reduces end-to-end coverage. Android cannot run Windows-only regression.
PR	PR #36092	Bounded condition wait after PostMessage for positive assertions.	WARNING INCONCLUSIVE (prior gate)	2 files	Original PR; simplest test-only change and code review found no concrete issue.

Cross-Pollination

Model	Round	New Ideas?	Details
gpt-5.5 / maui-expert-reviewer	2	Yes	Proposed a synthetic ActiveWindowTracker dispatch seam; captured as try-fix-4.
final assessment	2	No	Remaining options are trivial variations of polling, event signaling, synchronous dispatch, or test-seam abstraction split.

Exhausted: Yes
Selected Fix: PR's fix — The alternatives are plausible but none could be validated on the requested Android platform, and none is demonstrably better on code merits alone. Candidate 2 is the most deterministic but changes test fidelity; candidate 4 changes production internals for a test seam. The PR's bounded wait is the smallest localized test-only de-flake and preserves existing assertions.

---

📝 Recommended PR Title & Description

Assessment: ✏️ Recommend updating — the current description is detailed and accurate, but the title has CI-agent noise and does not match the [Platform] Component: What changed title formula.

Recommended title

[Windows] ActiveWindowTracker tests: Wait for posted messages before asserting

Recommended description

**Target branch:** net11.0
**Refs:** dotnet/maui#36043
**Attempt:** 1/5

## Summary

The `ActiveWindowTracker` Windows device tests intermittently fail with an empty-collection assertion — e.g. `SwitchingWindowsPostsToTheNewWindow` failing as `Assert.Contains() Failure ... Collection: []`. This is a **test-quality flake** (Step 4.7 bucket **b**), not a product defect.

## Root cause

`BaseWindowMessageManager_Tests.PostTestMessageAsync` posts an **asynchronous** window message via `PostMessage` and then waits a **fixed `await Task.Delay(100)`** before the test asserts the message was received. `PostMessage` only delivers once the message pump dispatches it; under CI load that can exceed 100 ms, so the positive assertions occasionally run before the message arrives and observe an empty collection.

Observed intermittently (green-on-retry) in 2 of the last ~10 `net11.0` `maui-pr-devicetests` builds (1473153, 1455596) — the classic flaky signal.

## Fix (de-flake — not a mute)

Replace the fixed-delay-then-assert pattern with a **bounded condition wait** (`WaitForMessageAsync`) for the three positive-assertion tests:

- `ActivatedBeforeStartRecievesMessage`
- `StartBeforeActivatedRecievesMessage`
- `SwitchingWindowsPostsToTheNewWindow`

The helper polls (yielding to the message pump) until the expected message is observed, up to a generous 5 s ceiling. It returns **as soon as** the message arrives, so there is no artificial slowdown, and if the message is **never** delivered (a genuine regression) the wait times out and the **unchanged** `Assert.Contains` still fails. Negative-assertion tests (`StoppedTrackerDoesNotTrack`, `SwitchingWindowsDoesNotPostToTheOldWindow`) intentionally keep the fixed delay, since absence cannot be polled for.

### Why this is not a mute

- No assertions removed or weakened — every original `Assert.Contains` / `Assert.DoesNotContain` is intact.
- No `[Retry]` / `[Repeat]` / `[ActiveIssue]` / `[Ignore]` / category exclusion / timeout-hiding change added.
- A real "message never delivered" regression still fails the test after the ceiling, so the change cannot hide a product bug.

## Validation

**Not runner-validated.** These are Windows-only device tests (`net11.0-windows`, `user32.dll` `PostMessage` + `Microsoft.UI.Xaml.Window`) and this fixer runs on a Linux runner with no Windows workload, so the `Essentials.DeviceTests` Windows path cannot be built or executed here. Please validate via a `maui-pr-devicetests` run on Windows (or run the Windows `ActiveWindowTracker` device tests locally) before merging.

## Files changed

- `src/Essentials/test/DeviceTests/Tests/Windows/BaseWindowMessageManager_Tests.cs` — add the `WaitForMessageAsync` condition-wait helper.
- `src/Essentials/test/DeviceTests/Tests/Windows/ActiveWindowTracker_Tests.cs` — await the condition before the three positive assertions.

---

🏁 Report — Final Recommendation

Comparative Candidate Report — PR #36092

Candidates Compared

Candidate	Approach	Regression status	Code-merit assessment	Rank
pr	Keep PostMessage; add bounded condition wait before positive Assert.Contains checks.	⚠️ Inconclusive, not failing. Windows-only tests could not be meaningfully verified on the requested Android/Linux environment.	Best balance: minimal test-only change, preserves queued message fidelity, preserves assertions, and directly waits for the assertion condition.	1
pr-plus-reviewer	Same as pr; expert reviewer produced no actionable changes.	⚠️ Inconclusive, not failing.	Equivalent to pr; no reviewer patch to distinguish it from the submitted PR.	1 (tie, no-op variant)
try-fix-1	Add a temporary TaskCompletionSource event signal completed by tracker.WindowMessage.	BLOCKED on platform; not regression-validated.	Precise and plausible, but adds per-test ceremony and cleanup risk without demonstrating a meaningful advantage over direct condition polling.	3
try-fix-2	Replace async PostMessage with synchronous SendMessage.	BLOCKED on platform; not regression-validated.	Most deterministic, but changes test fidelity by removing queued Win32 message delivery from the test path.	2
try-fix-3	Drain DispatcherQueue barrier turns after PostMessage.	BLOCKED on platform; not regression-validated.	Preserves queued delivery, but depends on ordering between Win32 posted messages and dispatcher queue callbacks; not clearly stronger than waiting for the observed condition.	4
try-fix-4	Add an internal synthetic dispatch seam and test tracker routing directly.	BLOCKED on platform; not regression-validated.	Deterministic at the abstraction boundary, but adds production test seam and reduces end-to-end coverage in these tests.	5

Regression-Test Ordering

No candidate has a validated passing Windows regression result in this environment. The gate is explicitly inconclusive, and the try-fix candidates are blocked because the requested platform was Android while the affected tests are Windows-only. No candidate is ranked as "passing"; no candidate is ranked lower due to a confirmed regression-test failure.

Winner

Winner: pr

The raw PR fix wins because it is the smallest localized de-flake that preserves the intended end-to-end PostMessage coverage and keeps all existing assertions intact. pr-plus-reviewer is code-identical because the expert reviewer produced no actionable findings, while the try-fix alternatives either add complexity, reduce fidelity, or introduce production test seams without validation evidence that they are better.

---

🧭 Next Steps — review latest findings

No alternative fix was selected for this run. Review the session findings and CI results before merging.

[PureWeen]

/azp run maui-pr-uitests

[PureWeen]

/azp run maui-pr-devicetests

[azure-pipelines]

Azure Pipelines:
Successfully started running 1 pipeline(s).

[azure-pipelines]

Azure Pipelines:
Successfully started running 1 pipeline(s).

[github-actions]

🎯 Target de-flake validated green — marking ready for review

SwitchingWindowsPostsToTheNewWindow is validated green on eee09e3 on the only platform it runs on (Windows), per the fresh maui-pr-devicetests run (build 1511744). The Helix testResults.xml reports result="Pass" for both packaged and unpackaged runs, and all five ActiveWindowTracker tests pass. The bounded condition-wait replaces the fixed 100 ms delay; no assertion was weakened.

The remaining red legs are unrelated pre-existing iOS/MacCatalyst Essentials FileSystem failures (CheckFileResultOpenReadAsyncMultipleTimes, OpenAppPackageFileAsync_Can_Load_File) — orthogonal to this Windows-only change, which is not even compiled on those platforms.

_{Automated by the CI Failure Fixer (net11.0). A maintainer should confirm before merge.}

Generated by CI Failure Fixer (net11.0) · 1.7K AIC · ⌖ 18.7 AIC · ⊞ 5.2K · ◷

[github-actions]

Target test SwitchingWindowsPostsToTheNewWindow validated green on eee09e3 on the only platform it runs on (Windows). Fresh maui-pr-devicetests build 1511744 Helix testResults.xml shows result="Pass" for both packaged and unpackaged runs, and all five ActiveWindowTracker tests pass. Remaining red legs are unrelated pre-existing iOS/MacCatalyst Essentials FileSystem failures, orthogonal to this Windows-only de-flake. Marking ready for maintainer review.

Generated by CI Failure Fixer (net11.0) · 1.7K AIC · ⌖ 18.7 AIC · ⊞ 5.2K

[azure-pipelines]

Azure Pipelines:
Successfully started running 1 pipeline(s).
There may be pipelines that require an authorized user to comment /azp run to run.

[kubaflo]

🔍 AI-generated review — automated multi-model consensus (Claude Opus 4.8 · GPT-5.5 · Gemini 3.1 Pro), independent reviews cross-pollinated and synthesized. Posted on behalf of @kubaflo.

Verdict: LGTM (code) — 3-model consensus, high confidence

What this changes: Adds a non-throwing poll helper WaitForMessageAsync(Func<bool>, timeoutMs = 5000, pollMs = 50) to BaseWindowMessageManager_Tests, and awaits it before the Assert.Contains(TEST_MESSAGE, …) checks in three ActiveWindowTracker_Tests cases (replacing reliance on a single fixed post-message delay).

Why it's a sound de-flake (verified):

• Failure detection preserved. The helper never throws on timeout — it just returns; the caller's Assert.Contains is still the correctness anchor and fails genuine non-delivery (after polling the full 5 s).
• No weakened assertion. In SwitchingWindowsPostsToTheNewWindow, TEST_MESSAGE is only ever posted to window2, and dispatch is per-HWND, so waiting on window2 cannot mask the DoesNotContain(window1…) check — which now runs later, giving strictly more time to catch a spurious entry.
• No data race. messages is a plain List<…>, but both the WndProc append (OnWindowMessage) and the Contains poll execute on the single WinUI UI thread (the window HWND is owned by it; await Task.Delay resumes on the same DispatcherQueueSynchronizationContext). Contains completes synchronously between await points, so no concurrent enumeration/mutation.
• Budget: ≤ 100 iterations; happy path resolves in ~100–150 ms, only genuine failures pay the 5 s.

CI status

• maui-pr-devicetests ● red, maui-pr-uitests ● red, maui-pr ● pending.
• Classification: failures confirmed PR-unrelated. The red device tests are CheckFileResultOpenReadAsyncMultipleTimes / OpenAppPackageFileAsync_Can_Load_File failing with System.NullReferenceException : Arg_NullReferenceException — the known dotnet/runtime#129813 net11 runtime-async miscompile in Stream.ReadAsync, not ActiveWindowTracker. This PR is a Windows-only test change and cannot affect those FileSystem tests.
• ℹ Final merge gating (green maui-pr) remains a human call.

Note for @kubaflo: LGTM — correct, race-free de-flake. Not formally approving (automated-loop policy).

[github-actions]

♻️ Attempt 1/10 — red is an unrelated CI flake, not caused by this PR (head eee09e3a4d0a6a80edd2be8fb1148a1aca7fc034)

The failing leg is Build macOS (Release), whose log shows Maui.Controls.Sample net11.0-ios erroneously compiling Platforms/Android sources (CS0246 Android not found, CS0101 duplicate AppDelegate, CS0115 override errors). This is the well-known MSBuild TFM-globbing infra corruption — it cannot be caused by this PR's Essentials/Windows ActiveWindowTracker fix, which touches no build glob or macOS/iOS asset.

A maintainer re-run (/azp run maui-pr) should clear the flake. No code change required.

Generated by CI Failure Fixer (net11.0) · 636 AIC · ⌖ 18.1 AIC · ⊞ 5.2K · ◷

[kubaflo]

LGTM — approving.

This PR replaces a fixed 100 ms delay with a bounded condition-wait in the Windows Essentials device tests (ActiveWindowTracker_Tests.cs, BaseWindowMessageManager_Tests.cs). The standing 3-model review (Opus 4.8 / GPT-5.5 / Gemini 3.1 Pro) reached LGTM.

CI-gate: the changed surface — maui-pr-devicetests (net11.0 Windows Helix Tests → Run DeviceTests Windows) passes, along with every Windows build/pack/unit/integration leg. The red legs are all unrelated:

• Build macOS (Release) fails inside Maui.Controls.Sample.csproj building for net11.0-ios (CS0246 'Android' not found, CS0115 OnAppearing/OnNavigatedTo no suitable method to override) — an orthogonal, transient net11.0 sample-app build breakage that a Windows-only Essentials-test change cannot cause.
• MacCatalyst/iOS CoreCLR device-test reds are the known net11 CoreCLR flake set (dotnet/runtime#129813); the remaining reds are non-Windows UI-test flakes.

🤖 AI-generated review (GitHub Copilot CLI, on behalf of @kubaflo).